Saturday, November 19, 2011 2:32 PM Jonathan Nieder : wrote
>forcemerge 627019 649304
>quit
>
>Hi Will,
>
>Will Set wrote:
>
>> [Subject: <processor.nocst=1> needed for 686-pae kernels]
>
>You haven't described the actual symptoms here. I assume that they
>are segfaults and hangs, as before?
Yes, the kernel still hangs early in boot. The difference being
the kernel hang without <processor.nocst=1> looks to me to be more
consistent now, than when I originally reported
the problem with the 2.6.39-rc7-686 kernel.
>
>> I first noticed this issue using 2.6.38-rc5-686 and 2.6.38-rc6-686.
>> Although, 2.6.38-1-686 worked out of the box, the regession showed again
>> using 2.6.39-rcX kernels,
>> and continues to show using all 686-pae kernels.
>> Adding <processor.nocst=1> to the kernel commandline works around the issue.
>>
>> Using 486 kernels is another workaround to this issue.
>
>Please send full "dmesg" output from booting up and reproducing the
>bug. Attaching "acpidump" output would be useful, too.
I'll also be away from my desk for a few days, and will followup,
with dmesg and acpidump, as soon as I return.
Best Regards,
Will
>
>> see: #627019 linux-image-2.6.39-rc7-686-pae: several kernel hangs ...
>
>Thanks for filing a separate bug so we can keep track of the
>information about each machine separately. I'm marking the two
>reports as related for easy reference.
>
>Thanks and hope that helps,
>Jonathan
